President's Address lv 
ferred in the form of rain to high land and mountainous regions, whence 
they very slowly soak again to the sea; nor are these areas of snow and 
rain symmetrical with respect to the axis of rotation. Further, these 
annual meteorological factors are very variable in extent, probably 
without regular law. From these causes there must arise annual and 
variable fluctuations in the ratio of the polar to the equatorial moments 
of inertia of the earth; and hence secondary variations not only in the 
amplitude of the angle between the instantaneous and the mean axes 
of the earth, but in the period of the rotation of the one axis about the 
other will be set up. Probably the combined effect of these variations 
will be to produce a curve expressing the total change of latitude which 
approximates in some degree to Chandler’s empiric law, but it is very 
improbable that the representation of the actual change of latitude can 
be thus expressed with the precision requisite for a definitive determ1- 
nation of the constant of aberration. ) 
By the method of observation recently proposed by Loewy, it is 
possible to determine the constant of aberration by a process which is 
entirely independent of the variation of latitude. Also, if the differ- 
ence of the meridian zenith distance of stars having nearly equal and 
opposite zenith distances be measured, these differences will be affected 
- in the opposite sense by aberration when observed in the evening and 
early morning. Consequently, if groups of stars symmetrically dis- 
tributed in right ascension are thus observed whenever they are visible 
in either of these circumstances throughout the year, it is possible to 
eliminate the constant of aberration from the observations entirely free 
from the effects of change of latitude. 
Airy’s experiments with the zenith telescope filled with water, com- 
bined with those of Michelson already mentioned, have set at rest most 
of the doubtful points in the theory of aberration, so far as they affect 
our problem. 
There is, however, a point about which we still know nothing, and 
that is the absolute velocity of translation of the solar system relative 
to the luminiferous ether which pervades space. If that velocity, as 
Villarceau has pointed out, has a sensible ratio, which we shall call A, 
to the velocity of light, then the observed value of the constant of 
aberration will be affected by a maximum quantity equal to Ax, the 
constant of aberration, and this will be multiplied by a further factor, 
never exceeding unity, depending upon the direction of the star whose 
aberration has been observed relative to the direction of the translation 
of the solar system through the luminiferous ether. It seems improb- 
able that this translation exceeds the known velocities of motion of 
the fixed stars with respect to each other, in which case the effect on 
the constant of aberration would be insignificant. But the question 
